JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 91, NO. A3, PAGES 3117–3129, MARCH 1, 1986
Aperture Plane Potential Control for Thermal Ion Measurements
R. C. OLSEN
Physics Department, University of Alabama in Huntsville
C. R. CHAPPELL
Space Science Laboratory
NASA Marshall Space Flight Center, Huntsville, Alabama
J. L. BURCH
Southwest Research Institute, San Antonio, Texas
Thermal plasma measurements from the Dynamics Explorer 1 spacecraft using an aperture bias in the outer plasmasphere
and over the polar cap show the existence of cold plasma (T ~1 eV) which may otherwise be hidden from the particle detector by
positive spacecraft potentials. Operations of the retarding ion mass spectrometer instrument with the external aperture plane biased
up to -8 V with respect to the spacecraft reveal the existence of low-energy (few eV) field-aligned flows over the polar cap.
Electrostatic models of the spacecraft indicate that if the spacecraft potential is more than a few volts positive, there can be a saddle
point, or an electrostatic barrier, inhibiting the effectiveness of the aperture bias. Comparison of the aperture bias experiments with
the electrostatic models shows that the aperture bias is successful in allowing the measurement of those field-aligned flows with
kinetic energy greater than the predicted barrier height but is less effective in aiding the measurement of the isotropic background.
The addition of electron measurements from the high-altitude plasma instrument allows the determination of the spacecraft potential
when it is greater than +5 V. This additional information allows us to determine that the dayside polar wind (reported by Sojka et al.,
1983) is supersonic, up to Mach 5.
1. INTRODUCTION
1.1. History
Plasma measurements in the magnetosphere have
always depended upon an accurate knowledge of the
satellite potential, particularly when the magnitude of the
satellite potential is comparable to the energy of the
measured particles. Measurement of the attracted species
generally only requires care in the data analysis, but
measurement of the repelled species requires the assumption
that the portion of the distribution which is measured is
characteristic of that at lower energies. Recent
measurements on geosynchronous satellites in eclipse show
that this is not always a valid assumption and that there can
be plasma distributions at low energies which cannot be
inferred from the higher-energy tail [Olsen, 1982].
Because of this problem, some form of potential control
is required, particularly for ion measurements in regions
where the satellite potential is positive, i.e., everywhere
outside the plasmasphere. An example of this is the S-302
experiment on GEOS 1 and GEOS 2, the suprathermal
plasma analyzers. These were electrostatic analyzers
mounted on 1.8-m booms, with a range of bias voltages
from +28 to -28 V. The 0.5- to 500-eV energy range was
designed to cover the thermal and suprathermal ion and
electron population of the outer plasmasphere and plasma
sheet. The biasing technique was successful, as long as the
bias voltage exceeded the spacecraft potential [Decreau et
al., 1978; Johnson and Wrenn, 1983; Olsen et al., 1983].
Copyright 1986 by the American Geophysical Union.
Paper number 4A8466. 0148-0227/86/004A-8466$05.00
For the retarding ion mass spectrometer (RIMS) on
Dynamics Explorer (DE 1), potential control was in the
form of an external aperture plane around the instrument
which could be biased with respect to the satellite. This is
not far removed from the original OGO 1 and 3 electron and
ion trap design, where the voltages could be applied to the
external grid, or aperture plane [Whipple and Parker,
.1969a,b; Parker and Whipple, 1970; Whipple et al., 1974].
Similar capabilities were incorporated on the mass
spectrometers on OGO 1, 2, 3, 4, and 6. The OGO 1
spectrometer, for example, had a guard ring which could be
set at 0, -5, -10, or -15 V. Much of the OGO 1 data was
taken at -15 V [Taylor et al., 1965]. The early OGO
experiences differed from those on later spacecraft because
the positive solar arrays and exposed solar array tabs
resulted in a negative satellite potential (-15 V at low
altitudes up to 0 V at high altitudes). Because of these
spacecraft potentials, a negative bias on the ion detectors
was not as useful as might have been anticipated.
Subsequent use of a biased guard ring on the Pioneer Venus
mass spectrometer has been infrequent, due to an apparent
lack of need for it in the Venus environment (H. A. Taylor,
private communication, 1984).
The analysis problem introduced by nonuniform
voltages such as those introduced by an aperture bias is that
an electrostatic barrier may be established which prevents
ions from reaching the aperture, even if it were energetically possible for them to do so in the absence of a barrier. In
particular, if the bias voltage is swept as on the OGO ion
traps, the analysis problem is made doubly complicated by
the variation in barrier height. The combinations of
potentials on DE 1 differ from those experienced on the
OGO satellites, but the analysis problem is similar. The
purpose of this paper is to illustrate the effect of an external
aperture plane on low-energy ion
JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 91, NO. A3, PAGES 3117–3129, MARCH 1, 1986
Magnetic field data are obtained from the DE 1 Goddard
Space Flight Center magnetometer [Farthing et al., 1981].
These data provide the pitch angle information needed to
interpret the particle angular distributions.
The high-altitude plasma instrument (HAPI) consists of
five sets of electrostatic analyzers arranged to cover most of
velocity space in one spin period, at energies from 5 eV to 32
keV. Ions and electrons are measured with 32% energy
resolution, 2.5° x 10° angular resolution, with a sample rate of
64 s-1, The geometric factor is 3.0 x 10-4 sr cm2 [Burch et al.,
1981].
Fig. 1. Orbit plot for DE 1, October 14, 1981 (day 287). The
orbit is plotted in SM coordinates, rotated to the local time of
each point, i.e. rotated into an L, λm plane.
measurements. An electrostatic model of the aperture bias
process is presented and compared to data taken in the
plasmasphere and polar cap.
1.2 Spacecraft and Instruments
The DE 1 satellite was launched into a polar orbit on August 3,
1981, with a perigee of 675 km altitude and an apogee of 4.65
RE. The 7.5-hour orbit began with apogee over the north pole,
with apogee moving toward the equator (in the midnight sector)
in the spring of 1982. The spacecraft spins in a reverse
cartwheel fashion, at 10 rpm with its spin axis perpendicular to
the orbit plane.
The retarding ion mass spectrometer (RIMS) is composed
of three sensor assemblies. The radial sensor head views
perpendicularly to the spin axis and provides pitch angle
distributions, while sensors on the ends of the space-craft view
parallel to the spin axis to complete the phase space coverage.
The radial sensor head has an angular resolution of ±10° in the
spin plane and ±55° perpendicular to the spin plane. The end
heads, or Z heads, have roughly conical apertures with 55° half
angles. The instrument covers the 1-32 AMU mass range with
two channels. The low channel covers the 1-8 AMU range, the
high channel 4-32 AMU. In this paper, we will show only H+
and He+ data using the low (1 AMU) and high mass (4 AMU)
channels in parallel. A retarding potential analyzer (RPA)
precedes the mass analysis, with a voltage sweep from 0 to 50
V. In addition to the RPA, there is an external aperture plane
which is normally grounded to the spacecraft body but can be
biased with respect to the spacecraft to values of -2, -4, and -8
V. The diameter of the aperture plane was 20 cm, substantially
smaller than the spacecraft dimensions of 1 m in height and 1.4
m in diameter. The retarding voltages are referenced to the
aperture plane, so that to the RIMS detector the effect of the
bias should be the same as a change in spacecraft potential,
aside from the effects of local fields [Chappell et al., 1981 ] .
2. OBSERVATIONS
The first orbit using aperture bias data was on October 14,
1981, illustrated in Figure 1, a projection of the orbit plane in
solar-magnetic coordinates. All of the data to be shown below
are from this orbit. The first example of aperture bias data is
from the plasmasphere, in order to illustrate the effect of the
bias in an isotropic plasma, with a spacecraft potential less than
0.5 V. Next, measurements of field-aligned ions on the nightside illustrate the results of aperture bias at spacecraft potentials
of a few volts. Finally, field-aligned ion measurements from the
dayside, at spacecraft potentials of 8 to 10 V, show evidence of
a barrier effect, and perhaps variations in the spacecraft
potential caused by the bias.
2.1 Observations in the Plasmasphere
The first data to be shown come deep in the plasma-sphere,
at 2 RE, near the magnetic equator (see Figure 1). The ambient
plasma density is about 5000 cm-3, and the spacecraft potential
is a few tenths of a volt negative. In this region, there should
not be any major effects due to local fields, since the Debye
length is less than 10 cm and thus less than the scale size of the
aperture plane. The hydrogen data illustrate subsonic flow
measurements, the helium data supersonic flow. Figures 2a and
2b show the retarding potential analyzer (RPA) voltage sweeps.
Data are taken in the ram direction and analyzed by using the
thin sheath approximation developed by Comfort et al. [1982,
1985]. This analysis technique is based on the general
formalism developed by Whipple et al. [1974]. Comfort et al.
calculated the integrated instrument response for a ram plasma
for the thin sheath case, obtaining an analytic form. The thin
sheath case is found to be close to the solution for an arbitrary
sheath, so long as the satellite potential is not large [Comfort et
al., 1985]. The RPA curves in Figures 2a and 2b shift in energy,
but otherwise, there are no apparent effects due to local fields,
as expected in this potential regime. The variation in flux with
the angle of attack (spin phase) depends upon the satellite
potential. The primary effect is a broadening of the spin curve,
as illustrated in Figures 2c and 2d, for hydrogen and helium,
respectively. In Figure 2c, the hydrogen data are plotted versus
the cosine of the ram angle. The fits to the data are based on
numerical simulations of the RIMS response to flowing
plasmas [Singh and Baugher, 1981; Comfort et al., 1985].
Analytical fits to the numerical simulations are used in the
fitting process. For an integral flux, 28' detector, the log of the
flux would be expected to show a linear dependence on cosine
theta, as would the log of the distribution function for a
differential detector
JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 91, NO. A3, PAGES 3117–3129, MARCH 1, 1986
Fig. 2. Isotropic plasma measurements at low altitudes made with the radial detector. (a) Hydrogen RPA curves take in the ram
direction (±10°) at 0, -2, and -4 V aperture bias. (b) Helium RPA curves taken in the ram direction. (c) hydrogen spin curve, 0 V
aperture bias, RPA voltages 0.0 to 0.5 V. (d) Helium spin curves, 0, -2, -4 V aperture bias, RPA voltages 0.0 to 0.5 V.
such as an electrostatic analyzer [Wrenn et al., 1984]. The
limited aperture, integral energy detector will not strictly
depend upon the cosine of the ram angle, but as illustrated in
Figure 2c, the deviation from a straight line is minimal until
entering the wake region (spin phase greater than 90°). Since it
is difficult to model the wake data anyway, this is not a serious
restriction. Just as with energy distributions, variations in
temperature (Mach number) cause the slope to vary. Variations
in spacecraft potential cause relatively smaller variations in
slope. Detector saturation effects are fairly severe for the
hydrogen data at this time, so we concentrate on the helium
data to illustrate the variations in the spin curves with aperture
potential.
Figure 2d shows the helium spin curves for aperture biases
of 0, -2, and -4 V. The peaks of the -2 V and much of the -4 V
data have been corrected for detector saturation effects,
beginning near the 80,000 count (6 MHz)level. It can be seen
that the peak flux increases, as might be expected, but more
importantly, the tails of the spin curves (the wake data) increase
by much larger percentages. This effect is also illustrated by
eclipse measurements reported by Olsen at al. [1985]. This
characteristic of the spin curves is important for the analysis of
the data taken at higher altitudes and the inference of barrier
effects which follows.
2.2. Observations in the Polar Cap
Continuing on the same day, we next present data taken at a
higher spacecraft potential, between +2 and +4 V. The upper
limit on the potential is determined from the electrostatic
analyzer data (HAPI). No spacecraft-generated photo-electrons
are observed by HAPI, indicating that the potential is below 4
V (the low-energy limit of the lowest channel). In the nightside
polar cap, the aperture bias
JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 91, NO. A3, PAGES 3117–3129, MARCH 1, 1986
Fig. 3. Isopotential contours for a model of the DE 1 satellite with biased RIMS radial aperture. The saddle points are
marked with crosses. Note that the contours plotted do not extend all the way to the aperture potentials (-2 and -6 V),
due to limits in the plot resolution.
operations revealed a high-velocity (10-20 km/s) fieldaligned hydrogen flow, in spite of the positive spacecraft
potential. Plate la shows the hydrogen data from the radial
detector in a spin-energy time spectrogram. The top panel
shows the RPA sweeps for the data taken along the
magnetic field line, the bottom curve the angular
distribution of the data taken at energies between 0.0 and
0.5 eV. The aperture bias is sequenced in a regular pattern
throughout this period, 0 V, -2 V, 0 V, -4 V, 0 V, -8 V, and
so on. The highest fluxes are seen at -8 V, as expected.
From 2034 and 2040 UT, the top panel shows green at
energies less than 1 eV, showing 5-10 counts per
accumulation in the field-aligned direction even at 0 V bias.
In addition to the field-aligned flow, a cold rammed plasma
appears at -4 and -8 V bias, as seen in the lower panel. The
detector view direction is plotted on the vertical axis of the
lower panel, with views in the direction of the spacecraft
motion (RAM direction) at 0°. The relationship to the
magnetic field directions is indicated by the solid and
dashed white lines running horizontally across the plot. The
solid line is for 180° pitch angle (flow from the north
geographic pole), while the dashed line is for 0° pitch angle.
These data have been interpreted by Nagai et al. [1984] as
the first high-altitude measurements of the supersonic polar
wind. The lower half of Plate 1 shows the He+ data for the
same time segment. A field-aligned helium population is
also found, as discussed (but not shown) by Nagai et al. The
helium flux is lower than the hydrogen flux by an order of
magnitude but is otherwise similar to the field-aligned
hydrogen flow.
The plasma wave instrument provides a different
perspective on this time period. Wave data can often be used
to obtain a plasma density, using features associated with
the plasma frequency. In this case, the wave data suggest
that this is an unusual day and that there may be an isotropic
background which is difficult to measure, even with the
aperture bias. The plasma wave data for this period have
been analyzed in concert with the other DE 1 experiments
by Gallagher et al. [1986]. They find that there may be a
background hydrogen population (non-field-aligned) with a
density near 10 cm-3. The question then is, do the aperture
bias operations reveal such a cold plasma? RPA curves
taken along the ram direction at -4 and -8 V bias (shown
below) show extremely low fluxes (10 counts or
JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 91, NO. A3, PAGES 3117–3129, MARCH 1, 1986
Fig. 4. Potential versus distance for the NASCAP DE 1/RIMS
modeL The barrier height is the 0.5-1.0 V maximum three to
six grid units from the aperture.
less per accumulation). RPA analysis without barrier effects
included produced densities of 1 cm-3 or less. In order to
further interpret these measurements, it is necessary to examine
the potential distribution around the satellite resulting from the
aperture bias operation.
3. MODELS
Interpretation of the plasma data requires a model for the
fields around the satellite similar to that constructed by
Whipple and Parker [1969a,b] for the planar geometry of OGO
1. Two models have been developed for this purpose and have
been compared to the data. These comparisons show qualitative
agreement between the data and models. The first model is the
NASA Charging Analyzer Program (NASCAP), a threedimensional numerical code designed to work on a 16 x 16 x 33
grid, solving Laplace's equation for a given potential
distribution and a specified geometry. Space charge can safely
be excluded for the 1-100 cm-3 density regime to be studied.
The addition of Debye shielding will only lower the barrier, so
we are effectively solving for the worst case. This code is
normally used to model spacecraft charging dynamics, solving
for the currents to and from the spacecraft, and iteratively
adjusting the spacecraft potentials and cur-rents. In this case,
only the potential solving portion of the code was utilized. This
code was previously used success-fully to model the differential
charging effects observed on ATS 6 [Olsen et al., 1981]. A
short cylinder was used to simulate the satellite, and a three cell
by three cell surface element was used to simulate the aperture
plane. The code resolution is essentially one cell.
The results for two sets of potentials are shown in Figure 3,
which displays isopotential contours in a cut perpendicular to
the spacecraft axis through the center of the detector. A saddle
point is formed in front of the aperture, with a magnitude below
1 V. The potential distribution along a nominal particle
trajectory radially away from (or toward) the detector is shown
in Figure 4. The potential maximum four grid units (0.4 m)
away from the detector is a barrier to low-energy ions.
A complete set of spacecraft and aperture potentials was
run with this model, with results shown in Figure 5, a plot of
the barrier height versus spacecraft potential for the different
aperture biases. The NASCAP predictions are shown in
combination with the results from an analytic model to be
described next.
The large size of the NASCAP code makes it awkward to
use, and a faster, more adaptable code is needed to study the
effects of aperture bias and angular effects. Also, by
compromising on the geometry of the model, higher resolution
is possible. Again, space charge is not included, and we solve
Laplace's equation. To accomplish this, the spacecraft was
modeled as a sphere, with the aperture modeled as a cap of 5°20° half angle. Setting the sphere at spacecraft potential and
biasing the cap to the aperture potential, it was possible to
expand the spherical potential distribution using Legendre
polynomials, in a multipole fit. The potential distribution away
from the sphere can then be accurately determined with this
spherical expansion, and these terms are easily evaluated.
The results from this model are illustrated in Figure 6, a
plot of the potential distribution for a cap of 10° half angle.
Again a barrier results. The barrier height obtained in this way
is also plotted in the smooth curves in Figure 5. The two sets of
plots agree in the monotonic rise in barrier height with
spacecraft potential, and the drop with increasing aperture bias.
The magnitudes are in rough agreement. Closer agreement
could be obtained with a different solid angle for the cap, but it
is felt that 10° accurately matches the ratio of areas on the
actual spacecraft. The difference seen in this plot is probably
largely due to the difference between cylindrical and spherical
geometry.
The effects such bathers should have on RIMS data are
illustrated in Figure 7. The barrier effect is incorporated into the
thin sheath model by assuming an average barrier across the
aperture which acts on the normal incident energy of the ions.
This is done by changing the low-energy
Fig. 5. Summary of the NASCAP barrier height calculations for
-2, -4 and -8 V biases for 1-5 V spacecraft potentials. NASCAP
model values are circled. The lines are the result of the
multipole sphere model
JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 91, NO. A3, PAGES 3117–3129, MARCH 1, 1986
cutoff (in the normal direction) to be the barrier height
(VBARR), when the sum of the RPA voltage and satellite
potential is less than the barrier height (VRPA + ri)S/C <
VBARR). The first-order correction should be sufficient, given
the limitations of the data. The left side of Figure 7 shows the
effects of a 1-V barrier on the measurement of a cold, isotropic
plasma. The data will appear to exhibit the effects of a -2 V
spacecraft potential, though the potential difference between the
aperture and ambient plasma is only 1 V. The right-hand side of
Figure 7 shows how a high-velocity flow can overcome barrier
effects, in anticipation of the polar wind measurements to be
shown.
3.1. Comparison of Data With Models
Fig. 6. Equipotential contours for a sphere at +2 V with a cap
biased -8 V with respect to the sphere (-6 V). A saddle point is
formed at +0.87 V.
3.1.1. Polar cap observations. The next step in the process is to
compare these models with particle data taken at high altitudes.
The field-aligned flows and the isotropic background illustrated
in Plate 1 are reasonably stable from 2033 to 2045 UT. The
spacecraft potential is apparently
Fig. 7. Barrier effects on RPA measurements are illustrated for a ram plasma, where the flow energy is less than the barrier height,
and a supersonic flow, where the flow energy is greater than the barrier height. The top figures show the ambient plasma
distribution, with the forbidden region shaded. The distribution function has shifted upward in energy when it reaches the negative
aperture, as shown in the middle panels. The RPA responses to such distributions are shown in the bottom panels.
JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 91, NO. A3, PAGES 3117–3129, MARCH 1, 1986
Plate 1. (a) RPA time (top panel) and spin time (bottom panel) spectrograms for the nightside polar cap hydrogen measurements.
The aperture bias goes through two cycles in the 12-min period shown here, -8 V, 0 V, -2 V, 0 V, -4 V, 0 V, -8 V, etc. The RPA
panel is restricted to data taken in the field-aligned direction (0 = 45°-90°). The spin panel is limited to RPA voltages from 0.0 to 0.5
V. The solid white line between 0 = 60° and 90' is the 180° pitch angle locus; the dashed white line between 0 = -90° and -120° is
the 0° pitch angle line. (b) As in Plate la, but for helium. The RPA panel is limited to 0 = 30°-75°.
JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 91, NO. A3, PAGES 3117–3129, MARCH 1, 1986
Fig. 8. Spin curves and RPA curves for the time period shown in Plate 1. (a) Spin curves for -8 V bias, 0.0-0.5 V RPA. The peak
near 90° spin phase can be fitted with a flowing Maxwellian: n = 2 cm-3,T = 0.2 eV, V = 20 km/s, = 0 V. The downward flow near 90° spin phase can be fitted with the same Maxwellian, but n = 0.02 cm-3. The isotropic background dominates the central portion of
the curve. (b) the -8 V bias data from Figure 8a, with 0, -2, and -4 V bias data. The upward flow persists even at 0 v bias, which can
be fitted with a flowing Maxwellian with n = 2 cm-3,T = 0.2 eV, V = 20 km/s, 0 = +3 V. (c) RPA data for the ram plasma (0 = -20°
to 0°). The dashed line indicates the portion of the fit truncated by the barrier. This corresponds to the forbidden region shown in
Figure 7. (d) RPA data for the upward field-aligned flow (0 = 55°-65°), are fitted with flowing Maxwellians.
+3 to +4 V, as deduced below, which suggests barrier
potentials of 1-2 V. These data should therefore be useful for
determining the effects of local fields on observations.
Figure 8c shows data taken in the "ram" direction, i.e.,
the flux perpendicular to the magnetic field line, the isotropic
background. The ram energy for hydrogen is negligible here
(less than 0.1 eV), so the knees in the RPA curves occur at
either the spacecraft potential or the energy-+ o5
difference between the aperture plane and the energy of the
saddle point (barrier). The "knee" for the -8 V bias data
occurs between 4 and 5 V; the "knee" for the -4 V bias data
occurs at about 1 V. These data would be consistent with a 3V spacecraft potential or a spacecraft potential of +4 V and a
barrier of slightly over 1 V. The latter situation is indicated
by the low densities (1 cm-3) resulting from attempts to
model the data without barrier effects. Also,
JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 91, NO. A3, PAGES 3117–3129, MARCH 1, 1986
Plate 2. (a) RIMS spectrogram for a 3-min period for the radial detector. The 6-s spin period produces the modulation of hydrogen
(lower panel) and helium (upper panel) fluxes. The aperture bias is cycled at 1-min intervals, here from -4 V to 0 V to -8 V. (b)
HAPI spectrogram for the same period as in Plate 2a. Electron fluxes in the top panel show high fluxes at low energies, nominally
locally generated photoelectrons trapped by a +10 V satellite potential.
JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 91, NO. A3, PAGES 3117–3129, MARCH 1, 1986
Fig. 9. Electron distribution function from HAPI. The locally
generated photoelectrons show a 1-eV spectrum up to 10 V,
followed by a warm (11 eV) ambient plasma distribution.
Ionospheric photoelectrons are found at 180° pitch angle.
These latter data were excluded from the least squares fit
(LSF) to the ambient distribution.
there should be a substantially larger difference between the
saturation levels of the -4 V and -8 V bias data, as indicated
by the dashed lines extending up from the fits to the data.
The nature of the problem is illustrated by Figure 7, which
shows how the energy of the distribution function varies
along the particle trajectories which reach the instrument.
Ambient ions with energies less than 1 eV cannot reach the
satellite. Those particles which do over-come the barrier then
fall down the potential hill into the detector, resulting in
integral flux measurements which are flat out to the energy
corresponding to the difference in potential between the
detector aperture and the barrier.
Figures 8a and 8b show the spin curves for the hydrogen
data taken at RPA voltages between 0 and 1 V. The increase
in flux with increasingly negative aperture bias voltages is
obvious. The field-aligned flux is visible at 0 V bias,
Fig. 10. RPA curves for the field-aligned hydrogen fluxes
illustrated in Plate 2a.
The RPA analysis results listed here were obtained with both
the thin sheath approximation [Comfort et al., 1984] and the
neutral planar approximation [Sojka et al., 1983] to facilitate
comparison with the previous analysis by Sojka et al. The
neutral planar model results are listed in parentheses. The
energy, E, in the fourth column is the flow energy. A
spacecraft potential of +8 to +10 V implies an ambient
electron density of 0.1-10.0 cm-3,
with a ram component becoming visible at -4 and -8 V bias.
At -8 V bias, there is a hint of a small flux of downward
flowing ions (0° pitch angle, -80° spin phase). One of the
major oddities of these spin curves is the lack of broadening
of the field-aligned measurements. Indeed, the only
difference between the 0 V bias and the -2, -4, and -8 V bias
along the field line is the normalization (i.e., density).
Modeling of the detector response to a change from 0 to -8 V
bias shows that the supersonic flow inferred from the 0 V
bias data [Nagai et al., 1984] would produce spin curves
which broaden substantially as the effective spacecraft
potential varies from +3 to -5 V. This interpretation ignores
the likelihood that there may have been an adiabatic decrease
in pitch angle width as the ions flow out of the ionospheric
source region. Such effects will
Fig. 11. Hydrogen spin curves for the same period shown in
Figure 10.
JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 91, NO. A3, PAGES 3117–3129, MARCH 1, 1986
produce some of the characteristics of these curves, since the
pitch angle distributions have widths of 10°-20°, i.e., less
than the detector width (20° x 110°). The ram plasma portion
of the spin curve is difficult to model with plasma parameters
inferred from the RPA curves unless barrier effects are
invoked. It appears that the bulk of the ram plasma is
excluded from the detector aperture by a barrier and tht the
character of the field-aligned flow measurements is
determined by a barrier of approximately IV. As the bias
increases from -4 to -8 V, the barrier height varies only
slightly, dropping a few tenths of a volt. This reduces the
effect of the 4-V change in detector potential substantially,
resulting in spin curves which are nearly unchanged, both for
the ram plasma and the field-aligned plasma.
Analysis for the RPA curves taken along the field line
turns out to be largely unaffected by the barrier, since the
supersonic flow has a bulk energy of 1-2 eV, sufficient to
overcome some of the +3 V spacecraft potential at 0 V bias,
or the 1 V barrier at -4 or -8 V bias. This interpretation is
illustrated by Figure 7, which shows how the energy of the
distribution function varies. The major effect is to reduce the
flux at 0 V RPA voltage, resulting in a slight underestimate
of the ion density. This is illustrated by Figure 8d, RPA
curves taken at each aperture bias level, with fits based on
the 0 V bias data.
3.1.2. Dayside plasma sheet observations. A slightly
different set of ambient plasma conditions is found later on
the dayside portion of the orbit, where the ambient plasma
density is lower. Plate 2 shows an energy time spectrogram
for 3 min of aperture bias data. This time period was shown
by Chappell et al. [1982, Plate 2d] and is the first reported
observation of the polar wind. These data were analyzed by
Sojka et al. [1983] .
Previous analysis of these data suggested relatively low
Mach numbers, perhaps even subsonic flow. A lack of
knowledge of the spacecraft potential was a major problem,
which has been resolved in this work by use of the electron
data from the HAPI experiment. The result is that the dayside polar wind, measured at L = 6, is highly supersonic, with
a Mach number of 4-7.
Electron data from the electrostatic analyzer are shown in
Plate 2b and Figure 9. Plate 2b is an energy time spectrogram
for the same time period as shown in Plate 2a. The locally
generated photoelectrons are visible as a yellow band from 5
to 10 eV. The electron distribution function is shown in
Figure 9. There is a break, or at least a change in slope, in the
distribution function at all pitch angles which separates the
locally generated distribution from the ambient population, at
the spacecraft potential. (See the appendix for a discussion of
this technique.) The break is at 10 eV, suggesting a
spacecraft potential of about 10 V. The ambient electron
population is characterized by a density of 2 cm-3 (assuming
a potential of +10 V) and a temperature of 11 eV.
Figure 10 shows two hydrogen RPA curves taken along
the field line at -4 and -8 V aperture bias. Using the spacecraft potential inferred from the electron data (+8 to +10 V),
a range of possible plasma parameters is obtained as listed in
Table 1. The electron data indicate that the ambient plasma
density is about 2 cm-3 , which is consistent with a +8 to +10
V potential. Additional information can be inferred from the
spin phase distribution shown in Figure 11. The lack of
displacement toward the ram direction shows that the fieldaligned flow velocity must be high compared to the satellite
velocity, i.e., of the order of 10-50 km/s [see Olsen et al.,
1985]. This combination of constraints argues for the Mach 5
to Mach 7 solutions. The -8 V bias data, previously thought
to indicate a +4 V spacecraft potential, can now be seen to be
the result of the large flow energy of the plasma (5 eV). The 4 V bias data are affected by a barrier which is +1 V with
respect to the aperture and +5 to +7 V with respect to the
ambient plasma. The -8 V data are associated with a 4- to 6V barrier which is just overcome by the high flow energy of
the hydrogen (see Figure 7).
This interpretation is supported by the spin curves,
shown in Figure 11. The data can be fitted with models
which match the 0.25 eV, 30 km/s interpretation of the RPA
curves, if a spacecraft potential near zero is assumed. Again,
however, the spin curves do not change shape as the aperture
bias is varied, indicating they are dominated by the 5- to 6-V
barrier. This interpretation again ignores the likelihood that
the spin curve is partly determined by the pitch angle
distribution.
4. SUMMARY AND CONCLUSIONS
Aperture bias data taken in the plasmasphere showed that
in the thin sheath regime, the detectors behaved as though
they were simply shifting in potential. This produced one
undesirable side effect, detector saturation. Aperture biases
larger than the plasma thermal energy degraded the effective
energy resolution of the instrument in this regime. Outside
the plasmasphere, with the space-craft potential at small
positive values, barrier effects reduced the effectiveness of
the bias, particularly in terms of the isotropic background
measurements. High Mach number flows were largely
undistorted in the ram direction, but spin curve variations are
dominated by the barrier height rather than the satellite
potential. The interpretation of the dayside polar wind is
shown to be highly super-sonic. This change from earlier
interpretation is based more on an accurate knowledge of the
satellite potential than on barrier effects.
Analysis of the aperture bias data sets shows that the
normal positive spacecraft potentials found in the outer
plasmasphere and beyond can be partially compensated by
means of the aperture bias. This method is limited by the
existence of barrier effects and a possible increase in spacecraft potential caused by the biasing. This latter effect would
not be as severe on a satellite with more conducting surface
area, and it remains clear that a conducting surface is
imperative for any scientific satellite. The conclusions of
Sojka et al. [1983] on requirements for future thermal plasma
measurements are reinforced by the analysis shown here.
Active potential control is needed for future satellites. This
can be accomplished by means of devices such as plasma
emitters which set the entire spacecraft to near zero volts
potential, alleviating the problem of nonzero potentials and
barrier effects [Olsen, 1981]. Electron detectors capable of
measuring the 0.5- to 50.0-eV energy range are needed to
complement the ion detectors. Such data are particularly
necessary in the absence of potential control. It is clearly
possible to measure the transport of thermal plasma into the
magnetosphere; the challenge is to do the job properly.
JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 91, NO. A3, PAGES 3117–3129, MARCH 1, 1986
APPENDIX
The technique for inferring positive satellite potential
uses the uniqueness of particle trajectories reaching the
detector at energies below and above the detector potential
(energy). Figure 12 illustrates the two classes of trajectories
which are allowed for a typical differential analyzer.
Trajectories for particles with energies less than the satellite
potential must return to the satellite surface, while those at
energies above the satellite potential reach the ambient
plasma. Unlike ambient plasma distributions, the two
distributions do not mix [Montgomery et al., 1973; Olsen,
1982]. This technique also depends on a lack of sensitivity of
the detector to EUV generated photoelectrons within the
instrument. The HAPI design is nearly immune to such
effects, as demonstrated by a near zero (background) count
rate when the satellite is in the plasma-sphere, and near zero
volts potential.
Alternate explanations of the electron distribution
function are possible. Given a photoelectron sheath around
the satellite, space charge effects could cause the formation
of a potential barrier to electrons leaving the satellite. This
would cause the phase space boundary to be at an energy
greater (with respect to the satellite) than the satellite
potential. The result would be an overestimate of the satellite
potential. Extensive thermalization of the local
photoelectrons in the sheath would blur the sharp boundary
in phase space, with uncertain effects on the potential
estimate.
Fig. 12. Electron trajectories and phase space boundaries for
a satellite charged to +10 V.
Acknowledgments. The authors would hie to express their
gratitude to the Data Systems Technology Program and the
Space Physics Analysis Network (SPAN) for providing
computer and networking facilities. Thanks to J. D. Menietti
for processing the HAPI data and to M. Sugiura for use of his
magnetic field data in the determination of the particle pitch
angles. The research at the University of Alabama in
Huntsville was supported by NASA contract NAS8-33982
and NSF grant ATM8-300426. The authors are indebted to
the engineering and science staff of the University of Texas
at Dallas and to the RIMS team at Marshall Space Flight
Center. We are grateful to the programming staff of the
Intergraph and Boeing corporations for assistance with the
data reduction software.
The Editor thanks E. C. Whipple and another referee for
their assistance in evaluating this paper.
REFERENCES
Burch, J. L., J. D. Winningham, V. A. Blevins, N. Eaker, W. C.
Gibson, and R. A. Hoffman, High-altitude plasma instrument
for Dynamics Explorer-A, Space Sci. Instrum., 5, 455-463,
1981.
Chappell, C. R., S. A. Fields, C. R. Baugher, J. H. Hoffman, W. B.
Hanson, W. W. Wright, H. D. Hammack, G. R. Carignan, and
A. F. Nagy, The retarding ion mass spectrometer on Dynamics
Explorer-A, Space Sci. Instrum., 5, 477-491, 1981.
Chappell, C. R., J. L. Green, J. F. E. Johnson, and J. H. Waite, Jr.,
Pitch angle variations in magnetospheric thermal plasma-initial
observations from Dynamics Explorer-1, Geophys. Res.
Lett., 9, 933-936, 1982.
Comfort, R. H., C. R. Baugher, and C. R. Chappell, Use of the thin
sheath approximation for obtaining ion temperatures from the
ISEE 1 limited aperture RPA, J. Geophys. Res., 87, 51095123, 1982.
Comfort, R. H., J. H. Waite, Jr., and C. R. Chappell, Thermal ion
temperatures from the retarding ion mass spectrometer on DE 1,
J. Geophys. Res., 90, 3475-3486, 1985.
Decreau, P. M. E., J. Etcheto, K. Knott, A. Pedersen, G. L. Wrenn,
and D. T. Young, Multi-experiment determination of plasma
density and temperature, Space Sci. Rev., 22, 633-645,
1978.
Farthing, W. H., M. Sugiura, B. G. Ledley, and L. J. Cahill,
Magnetic field observations on DE-A and -B, Space Sci.
Instrum.,5, 551-560, 1981.
Gallagher, D. L., J. D. Menietti, J. L. Burch, A. M. Persoon, J. H.
Waite, Jr., and C. R. Chappell, Evidence of high densities and
ion outflows in the polar cap curing the recovery phase, J.
Geophys. Res., 91, 3321-3327, 1986.
Johnson, J. F. E., and G. L. Wrenn, Ion measurements on GEOS
using a negatively biased sensor on a positively charged spacecraft, Proceedings of the 17th ESLAB Symposium on Spacecraft/Plasma Interactions and Their Influence on Field and
Particle Measurements, Noordwijk, The Netherlands, September 13-16, 1983, Eur. Space Agency Spec. Publ., ESA
SP-198, 63-68, 1983.
Montgomery, M. D., J. R. Asbridge, S. J. Bame, and E. W. Hones,
Low energy electron measurements and spacecraft potential:
VELA 5 and VELA 6, in Photon and Particle
Interactions with Surfaces in Space, edited by R. J. L.
Grard, pp. 247-261, D. Reidel, Hingham, Mass., 1973.
Nagai, T., J. H. Waite, Jr., J. L. Green, C. R. Chappell, R. C.
Olsen, and R. H. Comfort, First measurements of the
supersonic polar wind in the polar magnetosphere, Geophys.
Res. Lett., 11, 669-672, 1984.
Olsen, R. C., Modification of spacecraft potentials by plasma
emission, J. Spacecr. Rockets, 18, 462-469, 1981.
Olsen, R. C., The hidden ion population of the magnetosphere, J.
Geophys. Res., 87, 3481-3488, 1982.
JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 91, NO. A3, PAGES 3117–3129, MARCH 1, 1986
Olsen, R. C., C. E. McIlwain, and E. C. Whipple, Observations of
differential charging effects on ATS 6, J. Geophys. Res.,
86, 6809-6819,1981.
Olsen, R. C., P. M. E. Decreau, J. F. E. Johnson, G. L.
Wrenn, A. Pedersen, and K. Knott, Comparison of
thermal plasma observations on SCATHA and GEOS,
Proceedings of the 17th ESLAB Symposium on
Spacecraft/Plasma Interactions and Their Influence on
Field and Particle Measurements, Noordwijk, The
Netherlands, September 13-16, 1983, Eur. Space
Agency Spec. Publ. ESA SP-198, 57-61, 1983.
Olsen, R. C., C. R. Chappell, D. L. Gallagher, J. L. Green,
and D. A. Gurnett, The hidden ion population:
Revisited, J. Geophys. Res., 90, 12,121-12,132, 1985.
Parker, L. W., and E. C. Whipple, Theory of spacecraft
sheath structure, potential and velocity effects on ion
measurements by traps and mass spectrometers, J.
Geophys. Res., 75, 4720-4733, 1970.
Singh, N., and C. R. Baugher, Sheath effects on current
collection by particle detectors with narrow acceptance
angles, Space Sci. Instrum., 5, 295-305, 1981.
Sojka, J. J., R. W. Shunk, J. F. E. Johnson, J. H. Waite,
and C. R. Chappell, Characteristics of thermal and
suprathermal ions associated with the dayside plasma
trough as measured by the Dynamics Explorer
retarding ion mass spectrometer, J. Geophys. Res., 88,
7895-7911, 1983.
Taylor, H. A., H. C. Brinton, and C. R. Smith, Positive ion
composition in the magneto-ionosphere obtained from
OGO-A satellite, J. Geophys. Res., 70, 5769-5781,
1965.
Whipple, E. C., and L. W. Parker, Theory of an electron
trap on a charged spacecraft, J. Geophys. Res., 74,
2962-2971,1969a.
Whipple, E. C., and L. W. Parker, Effects of secondary
electron emission of electron trap measurements in the
magnetosphere and polar wind, J. Geophys. Res., 74,
5763-5774, 1969b.
Whipple, E. C., J. M. Warnock, and R. H. Winkler,
Effects of satellite potential on direct ion density
measurements through the plasmapause, J. Geophys.
Res., 79, 179-186, 1974.
Wrenn, G. L., J. J. Sojka, and J. F. E. Johnson, Thermal
protons in the morning magnetosphere: Filling and
heating near the equatorial plasmapause, Planet. Space
Sci., 32, 351-363, 1984.
J. L. Burch, Southwest Research Institute, P.O. Drawer 2851,
San Antonio, TX 78284.
C. R. Chappell, Space Science Laboratory, NASA Marshall
Space Flight Center, Huntsville, AL 35812.
R. C. Olsen, Physics Department, University of Alabama in
Huntsville, Huntsville, AL 35899.
(Received December 26, 1984; revised September 30, 1985;
accepted October 24, 1985.)